A Look at SBI FM's IVR Solution

by CXOtoday Staff    Jan 21, 2010

In a bid to reduce call wait times, SBI Funds Management (SBI FM) implements speech-recognition technology in their contact centers.

Overview of the company

SBI Funds Management Private Limited (SBI FM) is an asset management company, managing more than 100 schemes. It is a joint venture between State Bank of India and Soci t G n rale Asset Management, France and boasts of total assets under management (AUM) of more than Rs. 38,000 crore as on 30th October, 2009.

SBI Mutual Fund has an investor base of over 60 lakh, with a network of over 30 investor service centers, 57 investor service desks, 50 district organizers, and 60 business associates covering over 200 points of acceptance.

The challenge

SBI FM wanted to differentiate itself from its competition through outstanding products and services, backed up by state-of-the-art technology deployments. The firm has been experiencing a substantial growth in its customer base resulting in an exponential increase in calls to its contact centers.  The result was higher call wait times and increased customer impatience as agents tried to cope with the influx of calls.

In a bid to resolve this issue, SBI FM did an analysis of call types that might be better managed as self-service options within the interactive voice recognition (IVR) system and identified enquiries into the net asset value of funds as calls that could be effectively automated.

Implementing an open-dialogue call steering application in the IVR that would automate the NAV queries became an obvious solution - but that was not without its challenges due to the diverse range of schemes offered by SBI mutual fund, the number of different dialects spoken by its customer base and the way in which customers could reference the fund enquiry - i.e. by its full name, short form, or abbreviation of the fund name.

Thus creating an IVR system that could handle enquiries and route them appropriately and accurately was very important.

The solution

Following a detailed review of similar technologies, mid-2009, SBI MF commissioned Nuance, Datacraft, and Cisco to implement a speech recognition-based IVR solution for its contact centers using Nuance’s speech recognition engine "Nuance Open Speech Recognizer" (OSR) that allowed customers to access NAV details themselves.

The implementation was completed in around five months, and enables customers to speak to the IVR system to get their latest NAV balance.  This automation of routine, frequent enquiries now frees up the time of contact centre agents to address more complex queries.

Before the solution went live, the Datacraft team conducted test trials to fine-tune the application to adapt it to different dialects and make it more accurate. Datacraft experts tracked the trends in speaker dialects and the manner of requesting NAVs of mutual fund schemes.

Today the IVR with speech engine will prompt the caller for the fund name. The system will recognise the fund, retrieve the latest NAV value from back-end database system and play the NAV values to the caller. The system allows callers to request for NAV information of up to five funds. Customers navigate the system through speech-driven commands, which results in a pleasant user experience.


"We are constantly looking at innovating new schemes and introducing new solutions to enhance our business efficiency and win customer loyalty. This solution enables us to take our customer service to a new level," says Subhojit Roy, head (IT) of SBI FM.

As a result of the Nuance OSR deployment, SBI MF accrued a number of advantages like:

* Efficient customer service with reduced hold time
* Streamlined service, by eliminating the need to navigate lengthy IVR menus
* Personalized self-service experience for the customers.

SBI FM has realized significant business benefits since deploying the Nuance speech solution. The IVR now handles nearly 30 percent of calls previously managed by customer service representatives.  The implementation of this solution has also resulted in 20 percent more accuracy in the speech recognition application that will provide a concrete ROI in 15 months.

On the business front, it helped to reduce operational costs while improving service levels. It also gave agents more time to address complex queries while giving the company the ability to leverage existing infrastructure for greater volumes even as it reduces the cost per interaction and the call duration per customer.